MemoryblockMBS benchmark
The new 9.7pr14 plugins not just add a few more CUPS functions, we also got a nice Memoryblock vs. MemoryblockMBS benchmark project.As you see in the numbers on the left, the MBS class becomes very much faster compared to memoryblock if the size increases.
For smalls sizes the MBS class is slower due to the overhead of calling plugin code and the extra object plugin object being created. But on the raw memory allocation for the memoryblock, the plugin is faster. Instead of allocating a certain memory area and clearing it with zeros, the plugin orders directly cleared memory. As you see allocating a 1 GigaByte big memoryblock using REALbasic functions costs you 1.2 seconds while the plugin only needs 1.2 milliseconds. The times in the listbox are in microseconds.
